Live data from Hacker News

GitLab Pages

pages.gitlab.io

261–270 of 317 posts

Re: GitLab Pages

#261
post #72
post #54

Earlier quoted context omitted.

We're working on adding Cloudflare in front of GitLab Pages in https://gitlab.com/gitlab-com/operations/issues/75 We don't have a lot of experience yet with DDOS protection so there might be downtime as we learn that. As you can see in the issue we're working to implement protection before we need it but it might not be ready in time nor can it protect against everything.

Please do not use Cloudflare. Their 'protection' against privacy aware Tor users makes accessing sites through incredibly painful.

In all fairness, Tor is something of a cesspool and cloudflare is doing the best they can[1].

[1] https://blog.cloudflare.com/the-trouble-with-tor/

Re: GitLab Pages

#262
post #219
post #213

Earlier quoted context omitted.

I'm sad to see that. The subject was arguably right on topic. I get the feeling that it was taken down only because it was inconvenient for someone.

> I get the feeling that it was taken down only because it was inconvenient for someone It's a tempting feeling to get, and a contagious one, but no—the only inconvenience we care about is that of HN readers who don't want to read tedious flamewars tacked like burning barnacles onto the top comment. I'm kind of glad you said that, though, because usually the sinister implication goes the other way, Gitlab being a YC-…

How many up votes does that comment have? I want to know how many HNers are so offended by diversity initiatives that they will lobby companies to boycott github.

Re: GitLab Pages

#263

Earlier quoted context omitted.

Sensitive people enjoy being outraged.

Would you want to work at a company which has a bullet point "some of the biggest barriers to progress are white women"? That one struck me as particularly egregious. If you said that completely out of context, the average person is going to picture this coming out of some chauvinist white male. Ironically, it's coming out of people aiming for "diversity". Even in context, apparently the background is an article call…

I suggest you read the criticisms of 1st and 2nd wave feminism. Historically, mainstream feminism has not been supportive of women of color.

Re: GitLab Pages

#265
post #113

Earlier quoted context omitted.

GitHub Flavored Markdown is itself a rip of the markdown standard. Is it giving enough credit to markdown for GitHub to include the word "markdown" in the name? If so, then by extension it should be sufficient for GitLab to include "flavored markdown" in the name. ;)

The many "standards" of Markdown also now include CommonMark [1]. There is also some interesting history as to why it is not named "Standard Markdown". [1] http://commonmark.org/ [2] https://blog.codinghorror.com/standard-markdown-is-now-commo...

Man, as much as I admire and respect his intelligence, sometimes Gruber can be quite unreasonable.

Re: GitLab Pages

#266

Earlier quoted context omitted.

For what it's worth Cloudflare does not do anything different for Tor than for regular IPs. Since those are static pages with most likely not exactly questionable content the use of Tor is probably not all that important anyways that the total number of users that might be affected would be miniscule.

>the use of Tor is probably not all that important anyways Who are you to decide that my use of Tor is not important enough?

A site operator. It isn't censorship for someone to decide who they want to talk to.

CF is simply giving site owners levers to use if they so choose. While I am a strong supporter of Tor, privacy, and anonymity, CF has done nothing wrong here by allowing site operators to selectively choose who they wish to serve to.

Re: GitLab Pages

#267

Earlier quoted context omitted.

For what it's worth Cloudflare does not do anything different for Tor than for regular IPs. Since those are static pages with most likely not exactly questionable content the use of Tor is probably not all that important anyways that the total number of users that might be affected would be miniscule.

>the use of Tor is probably not all that important anyways Who are you to decide that my use of Tor is not important enough?

A site operator. It isn't censorship for someone to decide who they want to talk to.

CF is simply giving site owners levers to use if they so choose. While I am a strong supporter of Tor, privacy, and anonymity, CF has done nothing wrong here by allowing site operators to selectively choose who they wish to serve to.

Re: GitLab Pages

#268
post #8

Earlier quoted context omitted.

Clourfare provides free SSL - just proxy through them.

There are several problems with it: 1. the certificate is not generated by you (in the Free or the Pro plan) but by them and these certificates are for multiple domain names. So you end up sharing your certificate with other domains / websites of dubious nature that are also on CloudFlare's network. 2. CloudFlare only secures the connection between users and the CloudFlare network. It does not secure the connection b…

Why is #1 an issue?

Re: GitLab Pages

#269
post #208

Earlier quoted context omitted.

How would that work if Github pages don't support SSL?

GitHub supports SSL for *.github.io so you can have your custom domain with SSL on CloudFlare, and CloudFlare can proxy https://example.github.io/.. . on the other side. eg: End User https://example.com CloudFlare https://example.github.io

Rewriting the Host header in this way requires a non-free CloudFlare plan.

Re: GitLab Pages

#270

Earlier quoted context omitted.

It has integrated CI whereas GitHub requires using a third-party CI system, it's had Reaction Emoji (called Award Emoji for GitLab) for a while now, it has "Merge when build succeeds", CE is entirely FOSS, etc.

I'd honestly be really surprised if Github didn't end up acquiring one of those CI systems which they already integrate with.

There seems to be a healthy competition between various ci software that works with github. Why would github try to kill that and take on such a high touch area.

Its quite puzzling why gitlab went that route, they are basically discouraging ci systems from integrating with them. I think this is would turn out to be a bad decision for gitlab.

Post reply on HN